Year: 2025
Type: article
Abstract: This paper presents a novel approach to enhancing Data Leak Prevention (DLP) systems within enterprise environments through the application of deep learning techniques. Traditional DLP methods often struggle to effectively detect complex data exfiltration patterns and adapt to evolving threats, particularly when dealing with unstructured data formats and insider risks. To address these challenges, the proposed system leverages advanced deep learning models, including Convolutional Neural Networks (CNNs) and Bidirectional Long Short-Term Memory (BiLSTM) networks, to analyze structured and unstructured enterprise data in real-time. The system demonstrates high performance across key evaluation metrics, achieving an accuracy of 94.7%, precision of 93.1%, recall of 91.5%, F1-score of 92.3%, and AUC-ROC of 96.1%, significantly outperforming traditional DLP approaches. The integration of natural language processing and behavioral analytics enhances the system’s ability to detect sensitive data leaks with greater context awareness and minimal false positives. Additionally, the methodology incorporates real-world enterprise datasets and adheres to regulatory compliance standards, ensuring practical applicability and legal alignment. The findings underscore the potential of deep learning to transform enterprise data protection strategies by providing scalable, adaptive, and intelligent DLP solutions.
Authors: Greesham Anand, Shivaraj Yanamandram Kuppuraju, Sambhav Patil
Language: English
Subfield: Artificial Intelligence
Field: Computer Science
Domain: Physical Sciences
Open Access status: hybrid
OpenAlex ID: W4409893824